About me and this page
I hold a Dr.sc.ing. (since February, 2009) and I am a researcher at Institute of Applied Computer Systems, Riga Technical University, Latvia.
My research interests include (but aren't limited to):
- different kinds of predictive modelling, metamodelling (also known as surrogate modelling), feature selection and construction, model ensembling/averaging, as well as numerical and combinatorial optimization, heuristic search etc. (and practical applications of them);
- industrial design optimization;
- indoor positioning systems and location-based services;
- some time ago I was also interested in telemechanics, industrial network communication protocols, industrial automation, Object Linking and Embedding for Process Control (OPC), SCADA systems etc.
On this webpage you can find:
- a selection of my recent publications;
- a short description of a sparse polynomial regression modelling approach called Adaptive Basis Function Construction (ABFC) which I have developed;
- software tools (including open source) for regression modelling using Multivariate Adaptive Regression Splines (ARESLab toolbox for Matlab), M5' regression trees and model trees (M5PrimeLab toolbox for Matlab), Locally Weighted Polynomials, Radial Basis Functions, Polynomial Neural Networks induced by Group Method of Data Handling, Adaptive Basis Function Construction, and Shepard interpolation, including VariReg software tool for most of the above methods plus "full" polynomials, sparse polynomials, Kriging interpolation, and a number of model averaging/ensembling/combining methods.
- some software tools (including open source) for classification (mainly based on Logistic Regression);
- Stereoscopic Pong game in Java;
- some Wi-Fi based indoor positioning stuff;
- some datasets for regression and metamodelling;
- a list of useful links regarding predictive modelling software and other stuff.
